In [7]:
plt.figure(figsize=(16, 16))
plt.barh(sorted_df['state'],sorted_df['homes_sold'],color='coral')
plt.xlabel('Number Of Houses Sold')
plt.title('Single Family Residential')
Out[7]:
Text(0.5, 1.0, 'Single Family Residential')
Real estate has always played a dominant role in society and also has become expensive with the growing economy. From choosing the right city and the right time to invest to the demographic aspects that determine the value of the investment later in the years. Hence, knowing the market prices gives many an opportunity to negotiate and avoid overpaying.
How can data science be applied for real estate market analysis?
We will begin by importing several essential libraries, including pandas, numpy, plotly.express, and matplotlib. These libraries are indispensable for tasks related to data manipulation, data cleansing, and data visualization.
import pandas as pd
import numpy as np
import plotly.express as px
import matplotlib.pyplot as plt
Data provided by Redfin, a national real estate brokerage. Redfin has direct access to data from multiple local listing services. The data provided gives us the most reliable data on the housing market. This dataseet consists of 58 features. For this project, we have used the following 4 main features for our analysis:
state: This comprises the names of the states within the United States of America.
property_type: This categorizes properties into five major types: 'Townhouse', 'Single Family Residential', 'Multi-Family (2-4 Unit)', 'Condo/Co-op', and 'All Residential'.
median_sale_price: Within this, you will find the median sale prices of houses that were sold within the specified state.
homes_sold: This provides information regarding the number of houses that were sold.
Link for the Dataset : https://www.redfin.com/news/data-center/
df = pd.read_csv('state_market_tracker.tsv000',sep='\t')
df.head()
| period_begin | period_end | period_duration | region_type | region_type_id | table_id | is_seasonally_adjusted | region | city | state | ... | sold_above_list_yoy | price_drops | price_drops_mom | price_drops_yoy | off_market_in_two_weeks | off_market_in_two_weeks_mom | off_market_in_two_weeks_yoy | parent_metro_region | parent_metro_region_metro_code | last_updated | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | 2012-08-01 | 2012-08-31 | 30 | state | 4 | 33 | f | Iowa | NaN | Iowa | ... | 0.066667 | NaN | NaN | NaN | 0.000000 | 0.000000 | 0.000000 | Midwest Region | NaN | 2023-09-10 14:22:34 |
| 1 | 2021-02-01 | 2021-02-28 | 30 | state | 4 | 42 | f | Virginia | NaN | Virginia | ... | 0.227633 | 0.142857 | 0.064117 | 0.060213 | 0.305455 | -0.130319 | -0.295264 | South Region | NaN | 2023-09-10 14:22:34 |
| 2 | 2021-04-01 | 2021-04-30 | 30 | state | 4 | 49 | f | Minnesota | NaN | Minnesota | ... | 0.156430 | 0.138517 | -0.006933 | -0.036480 | 0.648354 | -0.013401 | 0.214179 | Midwest Region | NaN | 2023-09-10 14:22:34 |
| 3 | 2015-10-01 | 2015-10-31 | 30 | state | 4 | 2 | f | Missouri | NaN | Missouri | ... | 0.012556 | 0.217196 | 0.001813 | 0.217111 | 0.130876 | -0.002263 | 0.108191 | Midwest Region | NaN | 2023-09-10 14:22:34 |
| 4 | 2014-01-01 | 2014-01-31 | 30 | state | 4 | 6 | f | Nebraska | NaN | Nebraska | ... | 0.011383 | NaN | NaN | NaN | 0.378996 | 0.072626 | 0.046567 | Midwest Region | NaN | 2023-09-10 14:22:34 |
5 rows × 58 columns
df=df[['period_begin','state','state_code','property_type','median_sale_price','homes_sold']]
df.head()
| period_begin | state | state_code | property_type | median_sale_price | homes_sold | |
|---|---|---|---|---|---|---|
| 0 | 2012-08-01 | Iowa | IA | Multi-Family (2-4 Unit) | 58500 | 15 |
| 1 | 2021-02-01 | Virginia | VA | Multi-Family (2-4 Unit) | 225300 | 44 |
| 2 | 2021-04-01 | Minnesota | MN | All Residential | 308300 | 6959 |
| 3 | 2015-10-01 | Missouri | MO | Single Family Residential | 154600 | 4358 |
| 4 | 2014-01-01 | Nebraska | NE | All Residential | 135100 | 965 |
df=df[(df['period_begin']=='2022-01-01')]
df.head()
| period_begin | state | state_code | property_type | median_sale_price | homes_sold | |
|---|---|---|---|---|---|---|
| 150 | 2022-01-01 | Missouri | MO | Townhouse | 189100 | 44 |
| 288 | 2022-01-01 | Tennessee | TN | Single Family Residential | 334100 | 6339 |
| 505 | 2022-01-01 | Nebraska | NE | Multi-Family (2-4 Unit) | 479500 | 11 |
| 668 | 2022-01-01 | Missouri | MO | Condo/Co-op | 189200 | 338 |
| 772 | 2022-01-01 | Mississippi | MS | All Residential | 234600 | 1297 |
df1=df[['state','property_type','median_sale_price','homes_sold']]
df1.head()
| state | property_type | median_sale_price | homes_sold | |
|---|---|---|---|---|
| 150 | Missouri | Townhouse | 189100 | 44 |
| 288 | Tennessee | Single Family Residential | 334100 | 6339 |
| 505 | Nebraska | Multi-Family (2-4 Unit) | 479500 | 11 |
| 668 | Missouri | Condo/Co-op | 189200 | 338 |
| 772 | Mississippi | All Residential | 234600 | 1297 |
Before proceeding with data visualizations, we will concentrate exclusively on one property type: Single Family Residential, and arrange the data accordingly. This approach aims to enhance the clarity and simplicity of data interpretation.
df = df[(df['property_type']=='Single Family Residential')]
sorted_df = df.sort_values(by='homes_sold')
sorted_df.head()
| period_begin | state | state_code | property_type | median_sale_price | homes_sold | |
|---|---|---|---|---|---|---|
| 10373 | 2022-01-01 | Columbia | DC | Single Family Residential | 849900 | 82 |
| 8089 | 2022-01-01 | North Dakota | ND | Single Family Residential | 266600 | 179 |
| 31493 | 2022-01-01 | West Virginia | WV | Single Family Residential | 294600 | 287 |
| 29942 | 2022-01-01 | Vermont | VT | Single Family Residential | 347200 | 457 |
| 11253 | 2022-01-01 | Delaware | DE | Single Family Residential | 338900 | 461 |
Let us conduct a comparison of sales within the Single Family Residential property type across different states.
plt.figure(figsize=(16, 16))
plt.barh(sorted_df['state'],sorted_df['homes_sold'],color='coral')
plt.xlabel('Number Of Houses Sold')
plt.title('Single Family Residential')
Text(0.5, 1.0, 'Single Family Residential')
From the graph above we understand that most sales of Single Family Residential occurs in Florida and least in Columbia for the year 2022.
Similarly, we will examine the trends in median house prices for properties sold in various states.
fig = px.choropleth(df,
locations='state_code',
locationmode="USA-states",
scope="usa",
color='median_sale_price',
color_continuous_scale="Viridis_r",
)
fig.update_layout(
title_text = 'Median House Prices by State',
title_font_family="Times New Roman",
title_font_size = 22,
title_font_color="black",
title_x=0.45,
)
fig.show()
From this visual, we get a gist that the western part of the country is expensive in comparision to the eastern part. While Hawaii had the highest median sale price for each house sold.
Let us find which type of property is sold the most
grouped_property = df1.groupby('property_type')['homes_sold'].sum() #To group the property and find the total number of houses for each category.
grouped_property.plot(kind='pie', startangle=140)
<AxesSubplot: ylabel='homes_sold'>
From this graph we understand that All Residential type was most sold throughout the United State in year 2022
grouped_state = df1.groupby(['state'])['homes_sold'].sum()
print(grouped_state.sort_values(ascending=False))
state Florida 66171 California 49982 Texas 48103 New York 26513 North Carolina 25242 Illinois 21687 Georgia 21053 Pennsylvania 20832 Ohio 19355 Arizona 18833 Michigan 17761 New Jersey 17664 Virginia 17046 Tennessee 14122 Washington 13526 Indiana 11909 South Carolina 11595 Maryland 11583 Colorado 11259 Missouri 10804 Massachusetts 10577 Alabama 9280 Wisconsin 8966 Nevada 8504 Minnesota 8457 Oregon 8405 Oklahoma 7559 Connecticut 7224 Kentucky 7119 Louisiana 6625 Utah 5536 Arkansas 5402 Iowa 5042 Kansas 4579 Idaho 4461 Hawaii 2813 Nebraska 2672 Maine 2634 Mississippi 2561 New Hampshire 2483 Rhode Island 2244 New Mexico 1800 Montana 1771 Delaware 1372 Alaska 1231 Columbia 1212 Vermont 1139 South Dakota 1128 West Virginia 688 North Dakota 424 Name: homes_sold, dtype: int64
Florida has the highest number of sales
grouped = df1.groupby(['state', 'property_type'])['homes_sold'].sum()
print(grouped['Florida'])
property_type All Residential 33118 Condo/Co-op 7710 Multi-Family (2-4 Unit) 465 Single Family Residential 22246 Townhouse 2632 Name: homes_sold, dtype: int64
From this we can understand that All Residential brings the highest number of sales
All Residential SalesAR_grouped = grouped[grouped.index.get_level_values('property_type') == 'All Residential'].sort_values(ascending=False)
print(AR_grouped)
state property_type Florida All Residential 33118 California All Residential 25023 Texas All Residential 24090 New York All Residential 13270 North Carolina All Residential 12659 Illinois All Residential 10874 Georgia All Residential 10558 Pennsylvania All Residential 10431 Ohio All Residential 9708 Arizona All Residential 9420 Michigan All Residential 8905 New Jersey All Residential 8833 Virginia All Residential 8555 Tennessee All Residential 7086 Washington All Residential 6775 Indiana All Residential 5978 South Carolina All Residential 5808 Maryland All Residential 5797 Colorado All Residential 5638 Missouri All Residential 5420 Massachusetts All Residential 5292 Alabama All Residential 4644 Wisconsin All Residential 4496 Nevada All Residential 4254 Minnesota All Residential 4249 Oregon All Residential 4209 Oklahoma All Residential 3790 Connecticut All Residential 3612 Kentucky All Residential 3579 Louisiana All Residential 3321 Utah All Residential 2772 Arkansas All Residential 2714 Iowa All Residential 2536 Kansas All Residential 2298 Idaho All Residential 2233 Hawaii All Residential 1407 Nebraska All Residential 1342 Maine All Residential 1325 Mississippi All Residential 1297 New Hampshire All Residential 1243 Rhode Island All Residential 1122 New Mexico All Residential 911 Montana All Residential 897 Delaware All Residential 686 Alaska All Residential 616 Columbia All Residential 606 Vermont All Residential 578 South Dakota All Residential 570 West Virginia All Residential 346 North Dakota All Residential 219 Name: homes_sold, dtype: int64
Hence, Florida has the highest sales among all the states
We can derive a number of conclusions from the data analysis shown above. Particularly, Florida emerges as a significant market for real estate transactions. The observation that Florida's median home prices roughly match the national average serves as support for this judgement. Florida therefore displays a higher rate of real estate investment activity than other states. This pattern can be seen in a variety of property kinds, such as "All Residential" and "Single Family Residential" homes.
It is vital to understand that the dataset's values determine whether or not these findings hold true. Any changes to the data, such as the inclusion of new numbers or adjustments, may have an effect on the patterns and conclusions that have been detected.